It turns out the answer is yes, because it already is. AI is still relatively nascent in banking but is starting to gain purchase. A recent survey of 400 global banking executives, conducted by the Economist Intelligence Unit for banking platform provider Temenos, found that about 20 percent of them think AI will improve the user experience. And Gartner predicts that by 2020, a quarter of customer service operations will use virtual customer assistants.
But virtual customer assistants (such as Bank of America’s Erica, which the bank widely rolled out to its 25 million mobile customers in May) are just one of many ways banks use AI to enhance the customer experience. Banks can use artificial intelligence to identify fraud and cybersecurity risks and provide more personalized services.
Ultimately, AI for banks boils down to being able to make more intelligent decisions based on the data the bank already has about its customers, which in turn leads to better service.
How Banks Can Use AI to Help Customers
Much of the discussion around AI in banking has focused on whether or how much the technology will replace human jobs by automating routine tasks. However, banks should take a wider view of AI and realize it can help customers in numerous ways.
For example, AI can help detect fraud in real time since machine learning algorithms are likely able to spot patterns of suspicious behavior faster than a human can. The faster fraud is detected, the more quickly customers can be alerted.
Many banks are either already using AI in this way or are starting to explore it. Citibank, through its Citi Ventures arm, has made a strategic investment in Feedzai, a data science company focused on real-time fraud detection, according to Tech Emergence.
HSBC is teaming with a U.K.-based startup, Quantexa, to use AI to spot money laundering activity in customer networks, CIO Dive reports.
The same pattern recognition tools used to spot fraudulent activities can also be used to detect patterns of customer activity and recommend products and services based on them. For example, if the bank detects that a customer has made transactions at stores that sell baby products and also has been paying for visits to an OB-GYN, the bank might reasonably conclude that the customer will soon become a parent. Then, the bank can offer the client life insurance or a 529 plan for college savings, which the customer might not have known about.
And, of course, banks can use voice-activated AI for virtual assistants that can make transactions smoother and easier for customers. For example, Bank of America says Erica can search past transactions, such as checks or shopping activity; access key information, such as routing numbers or the closest ATM; schedule face-to-face meetings at a branch; view bills and schedule payments; lock and unlock debit cards; and transfer money between accounts or send money to friends with Zelle, a digital payments network.
AI for Banks Requires a Data Strategy
Ultimately, AI enables banks to pull structured and unstructured data from disparate sources, including external sources, and identify problems and solutions much faster than humans can. Platforms such as IBM Watson allow banks to do this with ease. For example, IBM notes, Watson adds a cybersecurity threat insight engine to the company’s cybersecurity platform, QRadar Security Analytics Platform, helping customers analyze threats up to 50 percent faster.
Banks can and should invest in AI where it makes sense for them to do so. But to take full advantage of such solutions, they need to develop a comprehensive data strategy that allows them to identify and link as many data sources as possible. Banks need to securely store and inventory their data and then determine how best to link it via AI software, so that they can get the most valuable insights and deliver the best service to customers.